The Bystander Effect in Multi-Agent Reasoning: Quantifying Cognitive Loafing in Collaborative Interactions
cs.MA, cs.AI
Submitted: 2026-05-11
Updated: 2026-09-22
License: http://creativecommons.org/licenses/by/4.0/
The gist: Multi-agent systems (MAS) assume that collaborating inherently improves Large Language Model (LLM) reasoning.
Terminology
Abstract
Multi-agent systems (MAS) assume that collaborating inherently improves Large Language Model (LLM) reasoning. We challenge this by demonstrating that simulated social pressure triggers an algorithmic ``Bystander Effect,'' inducing severe cognitive loafing. By evaluating 22,500 deterministic trajectories across 3 dataset contexts (GAIA, SWE-bench, Multi-Challenge) with 3 state-of-the-art (SOTA) models, we semantically audit internal reasoning traces against external outputs. We formalize the Interaction Depth Limit (D L), the exact plurality threshold where an agent's logical sovereignty collapses into social compliance. Crucially, we uncover the Sovereignty Gap: models frequently compute the correct derivation internally but suffer ``Alignment Hallucinations'' -- actively subjugating empirical evidence to sycophantically appease a simulated swarm. We prove that multi-agent social load is strictly non-commutative; the "brand" identity of the ``Lead Anchor'' auditor disproportionately dictates the swarm's integrity. These findings expose architectural vulnerabilities, proving that unstructured multi-agent topologies can degrade independent reasoning.
Sources
- Beyond the Attention Stability Boundary: Agentic Self-Synthesizing Reasoning Protocols
- The Inverse-Wisdom Law: Architectural Tribalism and the Consensus Paradox in Agentic Swarms
- Reasoning Models Generate Societies of Thought
- A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agents
Related papers
- Highway Congestion Reduction through Reinforcement Learning Based Eulerian Headway Control
- You Only Align Once: Propagating Cooperative Behaviors in Multi-Agent Systems through Seed Agents
- Deny Without Disabling: Authorization-Paired Evaluation and Control for Multi-Agent Systems
- MA-SAPO: Multi-Agent Reasoning for Score-Aware Prompt Optimization
- PeroMAS: A Multi-agent System of Perovskite Material Discovery
- StitchCUDA: An Automated Multi-Agents End-to-End GPU Programing Framework with Rubric-based Agentic Reinforcement Learning